arXiv research

New RL algorithms learn QSE from strategic feedbacks with sample efficiency.

Experimental economics has repeatedly demonstrated that the Nash equilibrium makes inaccurate predictions for a vast set of games. Instead, several alternative theoretical concepts predict behavior that is much more in tune with observed data, with the quantal response equilibrium as the most prominent example. However…

New framework recovers reward and rationality parameters from game behavior.

Unified framework for estimating reward functions in competitive games.

Study compares cryptocurrency and stock markets using statistical equilibrium models.

Model captures decision-making under bounded rationality with prior beliefs and market feedback.

QSD enhances deep network performance through biologically plausible dropout.

Although recent work in AI has made great progress in solving large, zero-sum, extensive-form games, the underlying assumption in most past work is that the parameters of the game itself are known to the agents. This paper deals with the relatively under-explored but equally important "inverse" setting, where the param…

The dual crises of the sub-prime mortgage crisis and the global financial crisis has prompted a call for explanations of non-equilibrium market dynamics. Recently a promising approach has been the use of agent based models (ABMs) to simulate aggregate market dynamics. A key aspect of these models is the endogenous emer…

Paper generalizes Hardy-Rogers maps for market equilibrium analysis in duopoly markets.

This paper extends a Kyle model to include price-responsive traders, revealing new dynamics and equilibria.

Paper studies zero-sum games with noisy observations and identifies equilibrium conditions.

New CGMD model predicts non-equilibrium processes better than existing methods.

In this paper, we propose an equilibrium pricing model in a dynamic multi-period stochastic framework with uncertain income streams. In an incomplete market, there exist two traded risky assets (e.g. stock/commodity and weather derivative) and a non-traded underlying (e.g. temperature). The risk preferences are of expo…

New algorithm for solving minimax problems over distributions converges to Nash equilibrium.

Study compares employers with and without anticipating strategic labor force responses.

Study best-response learning dynamics in zero-sum polymatrix games under full and minimal information settings.

We present a simple dynamic equilibrium model for an online exchange where both buyers and sellers arrive according to a exogenously defined stochastic process. The structure of this exchange is motivated by the limit order book mechanism used in stock markets. Both buyers and sellers are elastic in the price-quantity …

This paper simplifies complex game dynamics by using a recursive representation.

The large majority of risk-sharing transactions involve few agents, each of whom can heavily influence the structure and the prices of securities. This paper proposes a game where agents' strategic sets consist of all possible sharing securities and pricing kernels that are consistent with Arrow-Debreu sharing rules. F…

Paper introduces metrics for evaluating multi-agent policies using best response dynamics.

Investors' strategic trading affects asset prices, modeled as a game.

Study examines how market dynamics affect emissions trading prices and abatement efforts.

The paper analyzes a game where players must balance short-term and long-term interests, leading to cooperative or competitive outcomes.

New RL algorithms find SNE in Markov games with myopic followers.

Model analyzes trading frictions in cap-and-trade markets, showing how they interact to affect market effectiveness.

Informed traders strategically reveal noisier signals, making prices less responsive to public information.

Neural nets improve plasma equilibrium modeling for NSTX-U.

We address the challenge of designing optimal adversarial noise algorithms for settings where a learner has access to multiple classifiers. We demonstrate how this problem can be framed as finding strategies at equilibrium in a two-player, zero-sum game between a learner and an adversary. In doing so, we illustrate the…

Study efficient offline RL in Markov games with general models.

Develops variational framework for LQG risk-sensitive MFGs with major-minor interactions.

We give an elementary derivation of the Montgomery phase formula for the motion of an Euler top, using only basic facts about the Euler equation and parallel transport on the 2-sphere (whose holonomy is seen to be responsible for the geometric phase). We also give an approximate geometric interpretation of the geometri…

Policy mirror ascent achieves Nash equilibrium in mean field games without a population generative model.

Paper explores limits and possibilities of aligning LLMs with human preferences.

During the last two years, Europe has been facing a debt crisis, and Greece has been at its center. In response to the crisis, drastic actions have been taken, including the halving of Greek debt. Policy makers acted because interest rates for sovereign debt increased dramatically. High interest rates imply that defaul…

Perfect adaptation in systems is identified and tested using graphical tools.

New method finds invariant predictors to improve model robustness.

Study of portfolio management under relative performance concerns using mean field games.

The paper analyzes optimal stock position-building strategies in competitive markets.

We present the quantum model of Bertrand duopoly and study the entanglement behavior on the profit functions of the firms. Using the concept of optimal response of each firm to the price of the opponent, we found only one Nash equilibirum point for maximally entangled initial state. The very presence of quantum entangl…

Study shows climate change can cause a 'run on fossil fuels' affecting prices and production.

SPPO optimizes language model alignment by treating preferences as a game and achieving state-of-the-art performance.

The paper analyzes how investors' wealth can decline collectively under partial information.

A multi-layer deep Gaussian process (DGP) model is a hierarchical composition of GP models with a greater expressive power. Exact DGP inference is intractable, which has motivated the recent development of deterministic and stochastic approximation methods. Unfortunately, the deterministic approximation methods yield a…

Analyzing real data on international trade covering the time interval 1950-2000, we show that in each year over the analyzed period the network is a typical representative of the ensemble of maximally random weighted networks, whose directed connections (bilateral trade volumes) are only characterized by the product of…

Study time-inconsistent portfolio optimization for competitive agents with relative performance criteria.

Paper studies optimal tracking portfolio in mean field game of large fund competition.

Deep RL improves Diplomacy performance, outperforming previous methods.